Winner-Take-All Column Row Sampling for Memory Efficient Adaptation of Language Model.
Zirui LiuGuanchu WangShaochen (Henry) ZhongZhaozhuo XuDaochen ZhaRuixiang (Ryan) TangZhimeng Stephen JiangKaixiong ZhouVipin ChaudharyShuai XuXia HuPublished in: NeurIPS (2023)
Keyphrases
- language model
- memory efficient
- language modeling
- binary matrix
- speech recognition
- rows and columns
- probabilistic model
- n gram
- document retrieval
- test collection
- retrieval model
- information retrieval
- query expansion
- context sensitive
- language modelling
- language model for information retrieval
- language models for information retrieval
- vector space model
- translation model
- statistical language models
- mixture model
- smoothing methods
- bayesian networks
- document length
- statistical machine translation
- relevance model
- dependency structure
- dirichlet prior
- pseudo relevance feedback